THE RUNDOWN
Wednesday, 21 October 2026October 21: SpaceX, AI Safety, and Big Bets
SpaceX and Tesla might merge, Anthropic wants AI safety laws with real teeth, and OpenAI is moving into biotech. Busy Tuesday.
Elon Musk dropped a bombshell during SpaceX's Q3 earnings call on Tuesday: both the SpaceX and Tesla boards have formed independent committees to formally explore a merger between the two companies. That's not a rumour — that's the man himself confirming it. Meanwhile, Anthropic is trying to get Washington to take AI risk seriously, and OpenAI just bought itself a biotech startup. Let's get into it.
SpaceX and Tesla Eye a Merger
Musk said he thinks combining SpaceX and Tesla is a 'good idea' with 'a lot of merit,' pointing to overlap in energy systems, AI, and manufacturing. He was careful to note he's not sitting on either committee — both are made up of independent directors, which is the right move given his role at both companies. This would be one of the largest and most unusual corporate mergers in history if it goes through, and the market is already buzzing.
SpaceX Is Printing Money — Thanks to AI
SpaceX posted net income of $4.4 billion in Q3, and the engine behind that profit is AI — specifically its Starlink and compute infrastructure businesses, which are beating internal expectations. Musk also confirmed SpaceX will attempt to catch both Starship and its Super Heavy booster next month, aiming to recover the entire launch system for reuse. That'd be a first, and it would dramatically cut the cost of getting to orbit.
Anthropic's AI Safety Bill Has a Catch
Anthropic is pushing a federal AI safety proposal that sounds tough — it would flag models capable of helping cause 50 or more deaths. But read the fine print: the bill actually prevents federal agencies from blocking a model's release without the AI company's explicit permission. Critics say that's the opposite of safety regulation — it hands the keys to the companies being regulated. Worth watching how this shapes up in Congress.
OpenAI Moves Into Biotech
OpenAI has acquired Orchestra Bio, a life sciences startup, folding its team and technology into 'Rosalind' — OpenAI's AI model built for biological research. The Rosalind Workbench is designed to help scientists run and analyse experiments faster. It's a quiet but significant move that signals OpenAI is serious about healthcare and drug discovery, not just chatbots.
Anthropic Is Struggling to Find Compute
Here's an irony: Anthropic, one of the world's most talked-about AI labs, is reportedly having trouble locking down enough cloud computing power. The problem is that major cloud providers — AWS, Google, Microsoft — increasingly see Anthropic as a competitor and are less eager to prioritise its workloads. As AI infrastructure spending keeps climbing, this compute crunch is becoming a real strategic vulnerability. If you're following the AI industry closely, this one matters.
Gemini Is Coming to Your Car Dashboard
Google is rolling out Personal Intelligence for Gemini in Android Auto and cars with Google built-in. That means your car's AI assistant will soon be able to pull context from your Gmail, Calendar, Maps, and YouTube — so it can, say, remind you of a meeting while you're driving to it. Convenient, yes. Also a reminder of just how much data Google already has on you.
Netflix Has a Viewing Problem It's Not Talking About Loudly
Netflix's Q3 earnings report buries something interesting: average viewing time per subscriber is falling. The company is responding by making monetisation — think ads, live events, and games — its top priority rather than raw watch-time. It's a strategic pivot that's worth noting, especially as the streaming wars keep reshaping how entertainment gets made and sold in markets like ours.
Hugging Face CEO Calls Out Netflix's AI Documentary
Speaking of Netflix — Clem Delangue, co-founder of Hugging Face, is warning people not to take the new documentary 'AI Gone Rogue' at face value. He says it went 'full doomsday' and doesn't accurately represent the state of AI development or the people working in it. Given how much misinformation already circulates about AI, that's a flag worth raising before you share it.
PHP Warns of Emergency Security Patches
PHP has flagged that two emergency patches are coming very soon, following reports of a critical exploit called PHP2Shell that can remotely execute code on any server running PHP. Given that PHP still powers a massive chunk of the world's websites — including many across Africa — this is one to watch closely. If you manage websites or know someone who does, flag this and patch fast. Staying ahead of vulnerabilities like this one is non-negotiable.
The Next Xbox Could Cost $1,800
Microsoft is reportedly targeting a 2028 launch for its next-gen Xbox, codenamed Project Helix, with an eye-watering projected price tag of $1,200 to $1,800. The company is also apparently planning a gaming handheld for next year. At that price point, Microsoft is clearly going after a premium tier — whether the market follows is a different question entirely.
That's your Tuesday. The SpaceX-Tesla merger story alone is going to dominate tech conversation for weeks — keep your eye on what those independent committees actually recommend.
Comments
Get Weekly Tech Tips
Practical tech tips for Ghana, straight to your inbox. Free, every week.
No spam. Unsubscribe anytime.